Data
What real winning combinations look like: sums, splits and consecutive numbers
Winning tickets are not evenly spread across the play slip. The pattern is entirely a consequence of counting — and it matters for one reason only, which is not the one usually claimed.
Take every drawing since each game's current matrix began and measure the same five things about each one, and a very consistent portrait appears. None of it makes a combination more likely to be drawn. All of it explains why a random-looking ticket and a real winning ticket often look like different species.
The sum of the five white balls
The lowest possible Mega Millions sum is 1+2+3+4+5 = 15 and the highest is 66+67+68+69+70 = 340. In practice, sums pile up in the middle.
| Mega Millions | Powerball | |
|---|---|---|
| Drawings measured | 920 | 1,392 |
| Average sum | 174.3 | 176.9 |
| Standard deviation | 43.9 | 43.0 |
| Middle 80% of drawings | 118 – 229 | 120 – 232 |
| Lowest / highest observed | 48 / 326 | 52 / 299 |
There is nothing mystical here. There is exactly one way to make the minimum sum and exactly one way to make the maximum, but there are hundreds of thousands of ways to make a sum near the middle. Summing five draws from a flat range produces an approximately bell-shaped distribution — the central limit theorem doing its job on five samples. Every individual combination remains equally likely; the sums are not equally likely because they are not equally numerous.
Odd and even
How often each odd/even split appears, out of five white balls:
| Game | 0 odd | 1 odd | 2 odd | 3 odd | 4 odd | 5 odd |
|---|---|---|---|---|---|---|
| Mega Millions | 4.0% | 15.8% | 33.0% | 30.9% | 14.0% | 2.3% |
| Powerball | 2.6% | 14.6% | 31.0% | 32.0% | 17.0% | 2.9% |
The two middle columns — three-two splits in either direction — account for 63.9% of Mega Millions drawings and 63.0% of Powerball drawings. All-odd or all-even tickets do come up: 6.3% of the time in Mega Millions. Rare, but a long way from impossible.
Low and high
Splitting each pool in half — 1–35 against 36–70 for Mega Millions, 1–34 against 35–69 for Powerball:
| Game | 0 low | 1 low | 2 low | 3 low | 4 low | 5 low |
|---|---|---|---|---|---|---|
| Mega Millions | 1.8% | 14.6% | 30.3% | 34.2% | 15.8% | 3.3% |
| Powerball | 3.2% | 14.9% | 35.3% | 31.0% | 12.7% | 2.9% |
All five balls from the bottom half happened in 3.3% of Mega Millions drawings and 2.9% of Powerball drawings; all five from the top half in 1.8% and 3.2%. These are the drawings that make people say "that can't be random" — and they arrive at almost exactly the rate combinatorics predicts.
Spread and consecutive numbers
| Measure | Mega Millions | Powerball |
|---|---|---|
| Drawings with at least one consecutive pair | 25.8% | 26.4% |
| Drawings repeating a ball from the previous drawing | 32.8% | 31.9% |
| Balls landing in 4 or more different decades | 69.8% | 69.5% |
| Balls landing in 3 or fewer decades | 30.2% | 30.5% |
Roughly one drawing in four contains neighbouring numbers such as 34-35. Players avoid them instinctively, which means anyone who does play them is less likely to share a prize — the only sense in which the choice matters at all.
The birthday problem on a play slip
The most predictable human bias is the calendar. Dates cannot exceed 31, so tickets built from birthdays and anniversaries never use the upper two-thirds of the pool. How often does a real drawing produce five balls that would fit on a calendar?
| Game | Drawings | All five balls ≤ 31 | Expected by chance | Observed |
|---|---|---|---|---|
| Mega Millions | 920 | 21 | 12.9 | 2.28% |
| Powerball | 1,392 | 21 | 21.0 | 1.51% |
Powerball's count is essentially exact — 21 against an expected 21.0. Mega Millions has run a little hot, 21 against 12.9 expected, which a Poisson test puts at a 2.4% probability of happening by chance. That sounds impressive until you remember how many patterns are being tested at once: run twenty independent checks on random data and one of them will clear the 5% bar by construction. This is the multiple-comparisons trap that generates most "lottery pattern" claims.
The practical takeaway is not that calendar numbers are drawn less often — they are drawn exactly as often as anything else. It is that far more players choose them, so a calendar-only jackpot gets divided among more tickets. The famous illustration is the March 1989 Irish National Lottery drawing, and closer to home, every heavily-shared US jackpot has featured combinations rich in low numbers.
What to do with all this
Nothing about your chance of winning. Everything about what a win would be worth, and about recognising nonsense when you read it. The distributions above are what our Mega Millions and Powerball generators sample from when the "match historical shape" filter is on: instead of picking five numbers at random, they draw a real historical odd/even and low/high composition and fill it, so the lines they produce sit inside the distributions on this page rather than outside them.